The need for assessing impairment in functioning as an integral part of child mental health services research is discussed. Methodologic difficulties related to problems in case definition, the conceptual definition of impairment and social adaptation, and the assessment of the construct across cultures are also addressed. In addition, a critical review of existing child impairment measures with published psychometric properties is presented. The measures are divided into three types: measures of global impairment, domain-specific measures, and symptom-specific measures. The paper concludes that the choice of the measure to use in health services research needs to be determined according to the main goals of the study, the population in which the instrument is going to be used, and the interviewer's knowledge of the child. Other characteristics of the measures of impairment that the researcher should consider are also discussed.